Film Testing (WBM) Curves 320TXP in XTol 1:1 Too Contrasty—How Did I Mess Up?
I have just completed my first, Way Beyond Monochrome 2, film test, this morning. Since I have not used 320TXP (4x5) for a while (normally HP5+) I decided to test it, in my usual developer, XTol 1:1 20˚C. I use CombiPlan, with fairly intensive agitation: constant during first 30s (~15 full inversions), then 5 rapid full inversions taking a total of 6 seconds, every 30s, with a triple-rap-tap to dislodge bubbles. Incidentally, my sheets stay put, they never slip out with 320TXP, HP5+, or FP4 (I would get a rare slip with TMax 100). I did 5 sheets at a time.
Looking at the resulting curves (see the attached Excel spreadsheet), it seems I am getting too much contrast right from the beginning, even with short development times, eg. average gradient 0.59 at 4 min, and 0.92 at 16 min! It seems I cannot get N, not to mention N-1 this way... Clearly, I have done something wrong while testing.
I have a few suspects for the cause, and I wonder if those of you who have done this before, or who have other film testing experience, would be kind to suggest a clue and a fix, as I will re-run the test again in a while:
I think I have underexposed the whole thing by 1.5+ stops. WBM suggests metering at box speed, so I used ISO 320. After calculating the bellows factor (1.5 stops in this case), I was running against long exposure reciprocity (Irish summertime), and I went ahead with an uncorrected 1 s exposure at f/9, while I should have used a longer time. Further, my Nikkor runs 1 s as ~0.9 s. Can I blame my underexposing?
As I asked on another thread I was not sure how to photograph my Stouffer 31-step 4x5 transmission tablet. It was recommended to me that I should stick it to a window, as the test needed daylight (I photograph landscapes). That is what I did, but using 300mm Nikkor, with even a 52cm extension, all I managed to get is about a 70% reproduction, so the tablet on my negs is smaller than the real thing. I wonder if its overall density could get "squished" because of the smaller reproduction. Does the test require repro at 100% scale? If so, I am not sure how to do that with my field camera (Ebony SV45Te).
Through the window, I could clearly see distant objects (10-50m away). I placed a sheet of milky perspex about 8m away, angled towards the cloudy sky, outside in the garden, so I had a more even illumination behind the tablet. Even with the background well out of focus, I suspect there was a different illumination between the upper (1-16) and lower (16-32) tablet bars, as the tops of the bars 1-16 have less density than the bottoms of those bars on my 25 developed sheets (bars 16-31 are quite even). I collected the data consistently in the same spot, as close to the centre as possible, using a Heiland TRD-2. If this is the reason, then the through-the-window technique not going to work for me.
I worked fast, but this is Ireland, and the weather, cloudy with our normal drizzle, could have been slightly changing the lighting. I tried to compensate by shooting 5 sheets for each development time, and I have averaged readings within each development time series of 5 sheets, but I could not compensate for a gradual change in lighting across all the series. With hindsight, I should have grouped the sheets differently.
This box, like most of my film, has been through at least 2 airport hand-luggage x-rays, mandatory in Ireland and UK (no hand inspection allowed), but I doubt this would have other than the same overall effect.
I am agitating too much for those development times.
I am off to photograph in Wyoming for two weeks, tomorrow. I will use my 320TXP, and I plan to use the rule-of-thumb-rating of EI 200 for N, 250 for N+1 and 160 for N-1—suggestions? When I am back, I will run another test, hopefully helping me figure out the best dev times.
Please help me make the best choice, and thank you, for sharing your experience, for which I am always grateful.
I don't use your same testing method, so I cannot comment on your times. I get good prints with my equipment using a gradient around 0.75 and printing on multigrade paper.
I can say that if exposure is having an effect on the gradient there is a major problem in your method; like the values of your stepwedge are not evenly spaced or not evenly illuminated.
If you are not interested in calibrating your system for 'speed' then a contact print of the step wedge may be less problematic for you.
I know nothing of the WBM method so can't comment over points 1-5 but it seems to me that there may be substance in your concern over point 6. I have always understood that about 8-10 inversion in the first 30 secs( I prefer 8 inversions) were about right and about 2-3 inversions( I prefer 2 only) in 5 secs every thirty secs
5 inversions in 6 secs is better than a cocktail waiter can do
I'd certainly slow down and reduce both sets of inversions. I have no idea by how much the contrast will be reduced but it should help. I'd imagine that at 6 in 5 secs every thirty secs the Xtol must get quite bubbly. Mine did even at 3 inversions over 5 sec so I took it more gently.
Holy cow! Last time I got gradients like that the answer was obvious... I'd reached for the Dektol instead of the D-76
Bill, you seem to have taken a liking to film testing curves, judging by your earlier comments on CPorter's, and others' threads. I'd love to hear any suggestions you may have...
Originally Posted by pentaxuser
5 inversions in 6 secs is better than a cocktail waiter can do
I'd certainly slow down and reduce both sets of inversions. I have no idea by how much the contrast will be reduced but it should help. I'd imagine that at 6 in 5 secs every thirty secs the Xtol must get quite bubbly. Mine did even at 3 inversions over 5 sec so I took it more gently.
pentaxuser
Thanks. I get very even development, and no bubble-related issues, but I suppose this has to influence the gradient. I just wonder if that is the key reason for my numbers.
I don't use your same testing method, so I cannot comment on your times. I get good prints with my equipment using a gradient around 0.75 and printing on multigrade paper.
I can say that if exposure is having an effect on the gradient there is a major problem in your method; like the values of your stepwedge are not evenly spaced or not evenly illuminated.
If you are not interested in calibrating your system for 'speed' then a contact print of the step wedge may be less problematic for you.
I measured the original step-wedge—those numbers are also in the spreadsheet—and it seems quite even. I don't know if the 70% reproduction scale has any effect, though I would not think so.
How would I contact print a transmission step-wedge to a negative using natural light, with accurate timing? And also, how important is it that I do this test using natural light? Ralph, in his book, seems to suggest that it is important.
Thanks. I get very even development, and no bubble-related issues, but I suppose this has to influence the gradient. I just wonder if that is the key reason for my numbers.
No I didn't get any bubble related issues either it was just the fact that there were bubbles at the top of the Jobo tank when I removed the red cap to chek after each 30 secs that suggested to me that even 3 inversions was over vigorous if it made it froth as much as it did.
How would I contact print a transmission step-wedge to a negative using natural light, with accurate timing? And also, how important is it that I do this test using natural light? Ralph, in his book, seems to suggest that it is important.
This is how I do it, I put an 80A in there for daylight correction:
Other knowledgeable people (like PE) use an enlarger to make the contact exposure of the stepwedge on the film.
Unless you are testing for film speed, you don't need a precise exposure. You just need to get enough bands exposed to make a reasonable graph.
I just saw the part where you are taking off tomorrow... I think your rule of thumb for EI will work fine.
You can finish film testing when you get back. The meter readings and interpretation that leads you to Zone System processing notations (e.g., N-1, N and N+1) will not change if you change developer or times in developer later.
The only time this strategy (test later) failed me was when I rated some expired TRI-X at EI 320 when it later developed and proved to only be capable of approximately EI 64. So I had insufficient camera exposure to overcome fog. You will not have that problem.
I drew the graphs on paper and estimated the Contrast Index for each curve. Yes you could expose more next time (one stop), but that doesn't affect the information in the curves. CI by graph is somewhat subjective (the way I do it anyway) so a careful analysis could prove my numbers here are off by 0.10 on the steepest curve.
Adv Reply
Originally Posted by Bill Burk
